Prospective Value of PCR Amplification and Sequencing for Diagnosis and Typing of Old World Leishmania Infections in an Area of Nonendemicity

Abstract

We assessed the prospective value of PCR amplification of a repetitive sequence from Leishmania nuclear DNA and sequencing for the diagnosis and typing of Old World Leishmania infection in an area of nonendemicity. During this 42-month study, 29 of 168 consecutive samples were examined and classified as positive for Leishmania by direct examination and/or in vitro culture. This molecular approach showed excellent sensitivity (97%) and specificity (100%) compared to direct examination (86 and 100%, respectively) and in vitro culture (72 and 100%, respectively). Isoenzymatic and molecular typing allowed similar identification for 12 samples. Besides, PCR and subsequent sequencing of DNA products permitted the species identification of 14 samples for which parasite culture remained negative or did not allow isoenzymatic characterization, indicating the complementarity of parasitological and molecular tools.

The biological diagnosis of Leishmania infections usually relies on direct examination of smears after Giemsa staining and in vitro culture of clinical samples, i.e., bone marrow or blood in visceral leishmaniasis (VL) and dermal and mucous scrapings, aspirates, or biopsy material in cutaneous leishmaniasis (CL) and mucocutaneous leishmaniasis. These tools confirm the diagnosis when (i) amastigotes are detected in tissues or (ii) promastigotes are obtained in culture, thus allowing species identification by enzymatic characterization. However, these methods are tedious and time-consuming and require culture facilities and individual expertise, particularly in areas of nonendemicity where the number of cases detected is small. Few of the 50 to 100 samples that are collected in our laboratory each year for the diagnosis of Old World VL and CL are positive. This underlines the need for a trained microscopist, especially as in vitro culture is sometimes negative or contaminated by fungi. In case of positive culture, zymodeme analysis, which is performed only in reference laboratories (22), gives a definitive species identification within 2 or 3 months.

More recently, PCR amplification was also reported to be a useful method for the diagnosis of Leishmania infection. Several research groups designed and evaluated various PCR amplification targets (nuclear or minicircle kinetoplastic DNA, small-subunit rRNA, or miniexon RNA sequences) with primers defined according to their own endemic species (3, 4, 6, 7, 10-16, 18, 20, 21, 24). Typing methods for Leishmania species include multiplex PCR, randomly amplified polymorphic DNA, single-strand conformation polymorphism, restriction fragment length polymorphism, and sequence analysis (3, 4, 8, 9, 13, 17, 19, 23-25).

We prospectively evaluated the value of a PCR method for rapid diagnosis and typing of Old World VL or CL. We used PCR amplification of a repetitive sequence from Leishmania nuclear DNA as described by Minodier et al. (13), with further molecular typing by sequence analysis, for the diagnosis of Mediterranean VL. The sensitivity and specificity of PCR were compared with direct examination and parasite culture. Species identification by subsequent sequencing of PCR products was compared with zymodeme analysis of cultured promastigotes.

MATERIALS AND METHODS

One hundred sixty-eight consecutive samples obtained from patients originating from Europe or Africa were examined for Leishmania between June 1998 and December 2001. Twenty-eight samples (17%) were bone marrow aspirates, 46 samples (27%) were blood samples, and 94 samples (56%) were cutaneous samples (75 dermal scrapings and 19 biopsy specimens). Both conventional parasitological examination and molecular diagnosis were applied to each sample.

Parasitological diagnosis consisted of microscopic examination (magnification, ×1,000) of smears for amastigotes after Giemsa staining and in vitro culture in RPMI 1640 medium (Gibco, Cergy Pontoise, France) containing penicillin and streptomycin (BioMérieux, Marcy l'Etoile, France) and supplemented with 15% fetal calf serum (Gibco). When promastigotes were obtained in culture, isoenzymatic typing was performed by the French National Reference Center for Leishmaniasis (Montpellier, France), as previously described (22).

For molecular diagnosis, nucleic acid extraction was performed with QIAamp DNA mini kits (Qiagen, Courtaboeuf, France), according to the manufacturer's recommendations. PCR amplification was carried out with the modified primers T2 (5′-CGGCTTCGCACCATGCGGTG-3′) and B4 (5′-ACATCCCTGCCCACATACGC-3′) as previously described (13, 19). The amplification reaction mixture (50 μl) contained 10 μl of DNA extracted from a patient sample (either pure or diluted 1/10), 1× Applied Biosystems gold buffer, 2 mM MgCl₂, 200 μM dATP, 200 μM dCTP, 200 μM dGTP, 400 μM dUTP, 0.5 μM B4 primer, 0.5 μM T2 primer, 0.5 U of AmpErase uracil DNA glycosylase, and 1.25 U of AmpliTaq gold DNA polymerase. All reagents were purchased from Applied Biosystems except for the primers, which were obtained from Genset Oligos (Evry, France). Amplification was performed (pure and diluted 1/10 for each sample) on a Perkin Elmer GeneAmp PCR System 2400 thermocycler (Applied Biosystems) with a profile of 1 cycle of 5 min at 50°C; 1 cycle of 9 min at 95°C; 40 cycles of 30 s of denaturation at 95°C, 30 s of annealing at 56°C, and 30 s of extension at 72°C; and 1 cycle of 5 min of terminal extension at 72°C. The presence of amplification products was confirmed with 1% agarose gel electrophoresis analysis. For sequencing analysis, PCR products were purified on a Microcon 100 column (Amicon, Millipore, Molsheim, France). The two strands of amplified DNA were sequenced with the PCR primers and the BigDye Terminator kit (Applied Biosystems) on an automated sequencer (Applied Biosystems 377XL).

The genome and EMBL databases were consulted to obtain specific nucleotide information and, in particular, known polymorphisms of the available sequences. The nucleotide homologies of the sequenced products were studied with the BLASTN program (1) without the low-complexity filter (27).

The determination of consensus sequences and informative motifs was performed by using FASTA formatted sequences aligned with the MULTALIGN program (5) or the XALIGN program (26).

Nucleotide sequence accession number.

The repeat region sequence of Leishmania archibaldi has been submitted to GenBank under accession number AF344178 to enlarge the databases on this rare species. Original sequences of one strain of Leishmania major and one strain of Leishmania infantum obtained in this study have also been submitted to GenBank, under accession numbers AF421497 and AF421498, respectively, in order to improve the efficiency of sequence alignment.

RESULTS AND DISCUSSION

Leishmania infection was considered definite when the clinical diagnosis was confirmed by positive direct examination and/or in vitro culture. The sensitivity, specificity, and positive and negative predictive values were calculated for each diagnostic test. One hundred thirty-nine samples (83%) were negative by both parasitological and molecular tests.

Twenty-nine (17%) of the 168 samples were positive by direct examination, culture, or PCR, corresponding to 25 cases of CL and 4 cases of VL (Table 1). Clinical and biological findings allowed us to retrospectively exclude Leishmania infection for the 139 negative samples. No amplification was obtained with a few samples of New World Leishmania infections (data not shown).

TABLE 1.

Patient data and results^a

Patient no.	Sample origina	Country(ies) visited	Result by:		Isoenzymatic identification	PCR result	Molecular identification
			Direct examination	In vitro culture
1	DS	Senegal	+	+	L. major MON-26	+	L. major
2	CB	Algeria	+	−	ND	+	Failed
3	BM	France	+	+	L. infantum MON-24	+	L. infantum, L. donovani, L. archibaldi
4	CB	Mali	+	−	ND	+	L. major
5	DS	Algeria	−	+	L. major MON-25	−	ND
6	CB	Mauritania	+	−	ND	+	L. major
7	DS	Mali	+	−	ND	+	L. major
8	B	France	−	+	L. infantum MON-1	+	L. infantum, L. donovani, L. archibaldi
9	DS	Algeria	−	+	Failed	+	L. tropica
10	DS	Algeria	+	+	L. major MON-25	+	L. major
11	CB	Mali	+	−	ND	+	Failed
12	DS	Algeria	+	+	L. major MON-25	+	L. major
13	DS	Turkey	+	+	L. tropica	+	L. tropica
14	B	Sudan	+	+	L. archibaldi	+	L. infantum, L. donovani, L. archibaldi
15	DS	Senegal	+	+	L. major MON-74	+	L. major
16	DS	Senegal	+	+	Failed	+	L. major
17	DS	Tunisia	+	+	Failed	+	L. infantum, L. donovani, L. archibaldi
18	DS	Algeria	+	−	ND	+	L. major
19	DS	Turkey	+	+	Failed	+	L. infantum, L. donovani, L. archibaldi
20	DS	Mali	+	+	Failed	+	L. major
21	DS	Algeria	+	−	ND	+	L. major
22	DS	Sicilia, Baleares	+	+	L. infantum MON-34	+	L. infantum, L. donovani, L. archibaldi
23	DS	Senegal	+	+	L. major MON-74	+	L. major
24	DS	Algeria	+	+	Failed	+	L. infantum, L. donovani, L. archibaldi
25	DS	Mauritania	+	+	Failed	+	L. major
26	B	France	−	+	L. infantum MON-11	+	L. infantum, L. donovani, L. archibaldi
27	DS	Mauritania	+	+	Failed	+	L. major
28	DS	Senegal	+	−	ND	+	L. major
29	DS	Algeria	+	+	L. major MON-25	+	L. major
30-168	Various	Various	−	−	ND	−	ND

^aDS, dermal scraping; CB, cutaneous biopsy; BM, bone marrow; B, blood; ND, not determined; +, positive; −, negative.

Taken alone, direct examination had a sensitivity of 86% (25 of 29), whereas parasite culture had a sensitivity of 72% (21 of 29) (Table 2). PCR amplification yielded a specific product for 28 of the 29 samples from patients with proven leishmaniasis. Regarding patient 5, the false negativity of molecular amplification (pure and diluted 1/10) was attributed to PCR inhibitors since retrospective analysis of the sample allowed a positive result when diluted 1/50. Finally, this molecular method showed high sensitivity (97%) and excellent specificity (100%) (Table 2). Similar results have been obtained by some other research groups with various populations of patients (3, 4, 7, 8, 10-16, 18, 20, 21), particularly human immunodeficiency virus-infected patients, for whom tests with excellent sensitivity and specificity are necessary (2). Our results illustrate the complementarity of the parasitological and molecular methods for increasing the rapidity and sensitivity of diagnosis in areas of nonendemicity, where the small yearly number of positive samples can result in inefficient microscopic diagnosis.

TABLE 2.

Sensitivity, specificity, and positive and negative predictive values of direct examination, in vitro culture, and PCR for the diagnosis of Old World CL and VL

Diagnostic method	Result (% [no. positive/no. total]) for:
	Sensitivity	Specificity	Positive predictive value	Negative predictive value
Direct examination	86 (25/29)	100 (139/139)	100 (25/25)	96 (139/144)
In vitro culture	72 (21/29)	100 (139/139)	100 (21/21)	95 (139/147)
PCR	97 (28/29)	100 (139/139)	100 (28/28)	99 (139/140)

Regarding species identification by sequence analysis, the target sequence of PCR amplification is localized at the 5′ end of the previously described repetitive DNA sequence (L42476) of the sequenced clone LA6 from PstI-digested L. infantum MCAN/FR/73/LPMA/56 DNA (17). We have performed the alignments of the 31 repetitive Leishmania sequences available in the databases. The consensus sequence (Fig. 1) highlighted 3 polymorphisms (positions 65, 94, and 95) when the analyzed sequence was aligned with the easy-to-spot G₅ (5 consecutive guanines at positions 38 to 42). These polymorphisms easily categorize two species, L. major (D group) and Leishmania aethiopica (B group), and two species groups, L. infantum plus Leishmania donovani plus L. archibaldi (A group) and Leishmania tropica plus Leishmania killicki (C group) (Table 3). The analysis of other polymorphisms did not allow the differentiation of Leishmania species between the A and C groups. In addition to this typing method, we have determined 15 types of the polymorphic poly(A) regions (nucleotides 90 to 115 of the consensus sequence) (Table 4). However, analysis of this region did not allow us to discriminate one L. tropica strain from L. killicki (type 9) or one L. archibaldi strain from L. donovani and L infantum (types 1 to 2). Epidemiological data (clinical presentation and area of endemicity, etc.) are therefore of prime interest to interpret the molecular diagnosis. As a whole, sequence analysis of PCR products identified the species or the species group for 26 (90%) of the 29 patients. By comparison, in vitro culture of promastigotes was positive for 21 (72%) of the 29 samples. However, as isoenzymatic characterization failed for 9 strains, this method only yielded strain identification for 13 (45%) of the 29 patients. Most isoenzymatic characterization failures can be explained by insufficient growth of promastigotes in culture for isoenzymatic typing or by fungal contamination of the cultures. Results of the two typing methods were complementary, as shown in Table 1. In particular, PCR and sequence analysis allowed parasite identification in 14 samples (at the species level for 11 samples and at the species group level for 3 samples) that could not be cultured or identified by enzymatic methods and yielded the result within a few days, compared to several weeks for culture and subsequent isoenzymatic electrophoresis. Among the 12 patients for whom both isoenzymatic and molecular typing could be realized, 100% correlation between the two typing methods was observed. The reference method of identification by isoenzymatic characterization allows for these cases a precise species identification with a zymodeme determination that corresponds to a useful epidemiological tool.

FIG. 1.

Human pathogenic Old World Leishmania consensus sequence. The consensus sequence (bottom line) was aligned with the reference sequence of L. infantum L42476 (top line). *, unambiguous polymorphisms selected for the genetic diagnostic; —, ambiguous polymorphisms not selected. The G₅ region is underlined (position 38).

TABLE 3.

Genetic diagnostic polymorphisms at nucleotides 65, 94, and 95

Organism	Polymorphism(s) at nucleotide position(s)a:		Group
	65*	94*95*
L. archibaldi	G	ΔΔ	A
L. donovani	G	ΔΔ	A
L. infantum	G	ΔΔ	A
L. aethiopica	C	ΔΔ	B
L. killicki	C	CC	C
L. major	Δ	ΔΔ	D
L. tropica	C	CC	C

^a*, unambiguous polymorphisms selected for the genetic diagnostic; Δ and ΔΔ, simple and double deletions, respectively.

TABLE 4.

Genetic diagnostic polymorphisms of the polymorphic poly(A) region (nucleotides 90 to 115)

Type	Sequence (nucleotides 90 to 115)a	Reference strain(s)	Diagnosis
0	—GA———A——AA———AA———AGA—A——	Consensus
1	GGAA——A—GAAA—AAAAGCAGA—ACA	L. archibaldi AF344178, L. infantum L42481, L. infantum AF421498	L. archibaldi or L. infantum
2	GGAA——A—GAAA—AAAGGCAGA—AGA	L. archibaldi L42499, L. donovani L42475, L. donovani L42478	L. archibaldi or L. donovani
3	GGAG——A—GAAA—AAAAG—AGAGAGA	L. infantum L42476	L. infantum
4	GGAA——A—GAAA—AAAAG—AGA—AGA	L. infantum L42477, L. infantum L42479, L. infantum L42483, L. infantum L42486	L. infantum
5	CGAA——A—GAAA—AAAAG—AGA—AGA	L. infantum L42482	L. infantum
6	GGAA——A—GAAA—AAAAGCAGA—ACT	L. infantum L42480	L. infantum
7	—GAA——ATGAAA—AAAAGCAGA—AGA	L. aethiopica L42484	L. aethiopica
8	—GAA——ACGAAA—AAAAGCAGA—AGA	L. aethiopica L42485	L. aethiopica
9	GGAACCA——AAA—AAAAGCAGA—AGA	L. killicki L42500, L. tropica L42495	L. killicki or L. tropica
10	GGAA——A—GAAC—GAAAGTAGA—AGA	L. major L42501, L. major L42502, L. major L42503, L. major L42504	L. major
11	GGAA——A—GAAA—GAAAGCAGA—AGA	L. major L42505	L. major
12	GGAA——A—GAACGAAAAGTAGA—AGA	L. major AF421497	L. major
13	GGAACCA——AA——AAAAGCAGA—AGA	L. tropica L42487, L. tropica L42492	L. tropica
14	GGAACCA—AAAA—AAAA—CAGA—AGA	L. tropica L42488, L. tropica L42490, L. tropica L42491, L. tropica L42493, L. tropica L42494	L. tropica
15	GGAACCA——AAA—AAAA—CAGA—AGA	L. tropica L42489	L. tropica

^a—, ambiguous polymorphisms not selected.

These data make PCR amplification and DNA sequencing sensitive and efficient tools for rapid diagnosis and epidemiological studies of Old World leishmaniasis. However, remaining limitations incite us to use complementary parasitological and molecular tools to optimize the diagnosis.